S. Benferhat, D. Dubois, and H. Prade. How to infer from inconsistent beliefs without revising. In Proc. of the 14 International Joint Conference on Artificial Intelligence (IJCAI'95), pages 1449--1455, Montreal, 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... approach differs from other logic based approaches for handling inconsistent information such as belief revision theory (e.g. Gar88, DP98, KM91, LS98] knowledgebase merging (e.g. KP98, BKMS92] and logical inference with inconsistent information (e.g. MR70, Bre89, CRS93, BCD 93, BDP95] These proposals are too simplistic in certain respects for handling news reports. Each of them has one or more of the following weaknesses: 1) One dimensional preference ordering over sources of information for news reports we require finer grained preference orderings; 2) Primacy of ....

S Benferhat, D Dubois, and H Prade. How to infer from inconsistent beliefs without revising. In Proceedings of the 14th International Joint Conference on AI (IJCAI'95), 1995.

....The formalisms that we have described here for giving semantics to extended logic programs are paraconsistent in nature. i.e. they accept contradictions within the theory and try to cope with them. Another common approach to handle contradictions (sometimes called coherent or conservative [13,48]) rst detects and eliminates the inconsistent part of the theory. Then, when consistency is restored, some classical formalism is used for drawing plausible conclusions from the recovered data. In [31] for instance, clauses with negative literals in their heads are getting higher priorities ....

....already in the level of knowledge representation, since clauses for default rules have the form l Body; not l. Thus, in order to derive l, one has to verify rst that its complement, l, is not provable. Other coherent formalisms for managing inconsistent information are considered, e.g. in [2 4,7,13,18]. Belief revision The need to alter the set of conclusions according to an input that is frequently modi ed is not an unusual phenomenon in common sense reasoning in general and logic programming in particular. Thus, the plausibility of di erent formalisms in these areas is often determined by ....

S.Benferhat, D.Dubois and H.Prade, How to infer from inconsistent beliefs without revising ? in: Proc. International Joint Conference on Arti cial Intelligence (IJCAI'95), 1995, pp. 1449-1455.

....logic programs [40, 41] or bilattice based logics [5, 21, 33] together with non classical refutation procedures [20, 30, 40] that allow to detect inconsistent parts of a database and maintain them. A closely related topic is the problem of giving consistent query answers in inconsistent database [3, 10, 25]. The idea is to answer database queries in a consistent way without computing the repairs of the database. There are some other applications for integrating possibly con icting information and updating databases (e.g. LUPS [2] BReLS [31] RI [30] Subrahmanian s mediator of annotated databases ....

S.Benferhat, D.Dubois, H.Prade. How to infer from inconsistent beliefs without revising? Proc. IJCAI'95 , pp.1449-1455, 1995.

.... for dealing with this task are based on techniques of belief revision [20] methods of resolving contradictions by quantitative considerations (such as majority vote [21] or qualitative ones (e.g. de ning priorities on di erent sources of information or preferring certain data over another [2, 4, 5]) Other approaches are based on rewriting rules for representing the information in a speci c form [14] or use multiple valued semantics (e.g. annotated logic programs [28, 29] and bilattice based formalisms [12, 22] together with non classical refutation procedures [11, 19, 28] that allow to ....

S.Benferhat, D.Dubois, H.Prade. How to infer from inconsistent beliefs without revising? Proc. IJCAI'95 , 1449-1455, 1995.

....faulty components of malfunction devices, and database management systems that amalgamate distributed knowledge bases. 1 Motivation In this paper we introduce an algorithmic approach to revise inconsistent information and restore its consistency. This approach (sometimes called coherent [5], or conservative [15] considers contradictory data as useless, and uses only a consistent part of the original information for making inferences. To see the rationality behind this approach consider, for instance, the following set of propositional assertions: KB = fp; p; pq; r; rsg: ....

..... The set that is associated with is defined as follows: KB = f 2KB j ( t and A( I( KB) g: 1 Keeping this semantical correspondence to the original information is one of the main differences between the present formalism and some other formalisms for restoring consistency (see, e.g. [5, 6, 9]) The set KB corresponds to the (maximal) fragment of KB that can be interpreted in a consistent way by . Elimination of pieces of inadequate information in order to get a more robust representation of the intended knowledge is a common method in belief revision and argumentative ....

[Article contains additional citation context not shown here]

S.Benferhat, D.Dubois, H.Prade. How to infer from inconsistent beliefs without revising? Proc. IJCAI'95, pages 1449--1455, 1995.

....from all the possible world descriptions of KB. In the second part of the paper we extend the method mentioned above to cases in which the knowledge bases under consideration are prioritized. It is shown that like many other formalisms for reasoning with inconsistent and prioritized data (e.g. [1, 8, 9, 12, 17, 18]) our method is also nonmonotonic and paraconsistent [11] In addition, unlike some other formalisms like system Z [17] possibilistic logic [12] and default reasoning with conditional objects [10] the present approach is capable of managing ranked data without having the drowning problem [9, ....

.... 18] our method is also nonmonotonic and paraconsistent [11] In addition, unlike some other formalisms like system Z [17] possibilistic logic [12] and default reasoning with conditional objects [10] the present approach is capable of managing ranked data without having the drowning problem [9, 10]. 1 2 Background 2.1 The algebraic structure The approach that we consider here is based on Belnap s well known algebraic structure, FOUR, presented in [6, 7] This structure contains four elements: the two classical values, t and f , and two other values, and , that respectively denote ....

[Article contains additional citation context not shown here]

S.Benferhat, D.Dubois, H.Prade. How to infer from inconsistent beliefs without revising? Proc. IJCAI'95, pages 1449--1455, 1995.

.... 1994b; Nayak, 1994; Nebel, 1989; Nebel, 1991; Nebel, 1992; Rott, 1993 ] Further, similar approaches have been studied in the context of evaluating conditionals [ Ginsberg, 1986; Kratzer, 1981; Pollock, 1976; Veltman, 1976 ] hypothetical reasoning [ Rescher, 1964 ] and default reasoning [ Benferhat et al. 1995; Brewka, 1989; Poole, 1988; Reiter, 1987 ] The main idea in all these approaches is to start with a belief base and possibly a preference ordering over the formulae in the base or other means to express preferences and to generate a result by operations on the base. However, there are subtle ....

....the second case, we view the syntactic representation of a belief set as one ingredient for generating a belief revision operation on the generated belief set. Finally, it should be noted that in all syntax based revision schemes there is no arbitrary dependence on the syntactic representation [ Benferhat et al. 1995 ] a point we will return to in Section 10.2. 4.3 Analyzing the Computational Complexity of Base Revision Schemes For analyzing the computational complexity of base revision schemes, we could consider the problem of generating a belief base resulting from a revision scheme. However, in this ....

[Article contains additional citation context not shown here]

Salem Benferhat, Didier Dubois, and Henri Prade. How to infer from inconsistent beliefs without revising. In Proceedings of the 14th International Joint Conference on Artificial Intelligence (IJCAI-95), pages 1449-- 1455, Montreal, Canada, August 1995.

....expressive equivalence is a consequence of this result. ffl We formally clarify the connection between belief revision and non monotonic reasoning, in a particularly simple way which also throws light on the connection between consistence restoring and reasoning from inconsistency approaches [BDP95] ffl As a direct application of the above, we show that Poole s (syntax based) system of default reasoning and Shoham s preferential semantics for non monotonic reasoning are also expressively equivalent, in that they can represent the same set of non monotonic consequence relations. KEYWORDS: ....

.... revision is postulated as an attempt to restore consistency (again as in the AGM approach) or instead as a way to cope with inconsistency in the reasoning processes, allowing for meaningful reasoning from possibly inconsistent information (without attempting to remove the inconsistency, see e.g. BDP95] In this paper we present a number of results on the expressive equivalence of the various approaches. Let us define what we mean by this. A revision operator is a function from current beliefs and new information into revised beliefs . For concreteness we may assume that these three items ....

[Article contains additional citation context not shown here]

Salem Benferhat, Didier Dubois, and Henri Prade. How to infer from inconsistent beliefs without revising. In Proceedings of the Fourteenth International Joint Conference on Artificial Intelligence, pages 1449--1455, 1995.

....the most credible belief(s) The good with the highest priority in G . will be selected in S5 as the preferred good. In case of ties, the revised cognitive state may be based on either one of the goods with the same highest priority (randomly selected) or on their intersection (see [8] and [15]) This latter case means rejecting all the conflicting but equally credible information items. The result is not a good (it is not maximally consistent) and thus implies rejecting more assumptions than necessary to restore consistency. 3. Forming a collective opinion In order to improve their ....

Benferhat S., Dubois D. and Prade H., How to infer from inconsistent beliefs without revising?, in Proc. of the 14th Inter. Joint Conf. on Artificial Intelligence, pp. 1449-1455, 1995.

....systems in the taxonomy in which there can be misperception in the form of erroneous observations. In (Coradeschi 1995) is explained in details how our work relate with Sandewall s approach to the representation of knowledge. Our approach can be considered a coherence based approach as defined in (Benferhat, Dubois, Prade 1995), that is an approach where the aim is revising the knowledge base and restoring consistency. In particular it can be considered near Rescher and Mannor approach to reasoning with inconsistency (Rescher Mannor 1970) The problem of unreliable observations is related to the problem of belief ....

Benferhat, S., Dubois, D., and Prade, A. 1995. How to infer from inconsistent belief without revising? In Proceedings the 14th International Joint Conference on Artificial Intelligence, 1449--1455. Montreal, Canada: International Joint Conferences on Artificial Intelligence, Inc.

....common when revising inconsistent knowledge bases; If some formulae are more certain than others, one would probably like to reject the least certain first. Many different approaches for resolving conflicts in prioritized knowledge bases have been proposed in the literature (see, e.g. [5, 6, 8, 9, 11, 12]) A lot of these methods draw conclusions based on maximal consistent subsets of the knowledge base under consideration (see, e.g. 6] for a survey) However, the semantics of the maximal consistent sets might not correspond to that of the original knowledge base. For example, none of the ....

....certain first. Many different approaches for resolving conflicts in prioritized knowledge bases have been proposed in the literature (see, e.g. 5, 6, 8, 9, 11, 12] A lot of these methods draw conclusions based on maximal consistent subsets of the knowledge base under consideration (see, e.g. [6] for a survey) However, the semantics of the maximal consistent sets might not correspond to that of the original knowledge base. For example, none of the maximal consistent subsets of the simplest inconsistent knowledge base fp; pg reflects its intended meaning. Moreover, each maximal ....

[Article contains additional citation context not shown here]

S.Benferhat, D.Dubois, and H.Prade. How to infer from inconsistent beliefs without revising? Proc. 14th Int. Joint Conf. on AI, pages 1449--1455, 1995.

....from all the possible world descriptions of KB. In the second part of the paper we extend the method described above to cases in which the knowledge bases under consideration are prioritized. It is shown that like many other formalisms for reasoning with inconsistent and prioritized data (e.g. [1, 8, 9, 12, 18, 19, 20]) our method is also nonmonotonic and paraconsistent [11] In addition, unlike some other formalisms like system Z [18] possibilistic logic [12] and default reasoning with conditional objects [10] the present approach is capable of managing ranked data without having the drowning problem [9, ....

.... 20] our method is also nonmonotonic and paraconsistent [11] In addition, unlike some other formalisms like system Z [18] possibilistic logic [12] and default reasoning with conditional objects [10] the present approach is capable of managing ranked data without having the drowning problem [9, 10]. 2 Background 2.1 The algebraic structure The approach that we consider here is based on Belnap s well known algebraic structure, FOUR (Figure 1) presented in [6, 7] This structure contains four elements: the two classical values, t and f , and two other values, and , that respectively ....

[Article contains additional citation context not shown here]

S.Benferhat, D.Dubois, H.Prade. How to infer from inconsistent beliefs without revising? Proc. 14th Int. Joint Conf. on AI (IJCAI'95), pages 1449--1455, 1995.

....what is incorrect. We propose a solution where the agent, in case it finds a contradiction, construct several alternative descriptions of the world, each one containing a consistent subset of the observations received. Our approach can be considered a coherence based approach as defined in [BDP95], that is an approach where the aim is revising the knowledge base and restoring consistency. In particular it can be considered near Rescher and Mannor approach to reasoning with inconsistency [RM70] The problem of unreliable observations is related to the problem of belief revision and in ....

S. Benferhat, D. Dubois, and H. Prade. How to infer from inconsistent belief without revising? In Proc. IJCAI'95, pages 1449--1455. 1995.

....most credible belief(s) 2.4 S5 The good with the highest priority in G . will be selected in S5 as the preferred good. In case of ties, the revised cognitive state may be based on either one of the goods with the same highest priority (randomly selected) or on their intersection (see [8] and [15]) This latter case means rejecting all the conflicting but equally credible information items. The result is not a good (it is not maximally consistent) and thus implies rejecting more assumptions than necessary to restore consistency. 3 The Simulation Experiment The task given to the group is ....

Benferhat S., Dubois D. and Prade H., How to infer from inconsistent beliefs without revising?, in Proc. of the 14th Inter. Joint Conf. on Artificial Intelligence, pp. 1449-1455, 1995.

....literature. In (Cayrol 1995) an argument of D is a pair (H, h) where h is a formula of L (conclusion) and H a subbase of D (support) iff (i) F H is consistent, ii) F H h, and (iii) H is minimal (no strict subset of H satisfies (ii) A similar notion of argument has been also given in (Benferhat Dubois Prade 1995). According to them, a consistent subbase A of L is said to be an argument to a rank i for a formula f if it satisfies the following conditions: i) A 1 they may also contain subissues; due to space limitations, we don t describe them in detail here. f, ii) a A, A a f, and ....

Benferhat, S., Dubois, D., Prade, H. 1995. How to infer from inconsistent beliefs without revising? In Proceedings of the 14th IJCAI, 1449-1455.

....have been proposed in the literature [36] in order to address various application areas [4] For instance, in [11] an argument consists of a support base, that may contain formulas which speak for or against a certain position, and a conclusion. A similar notion of argument has been given in [5]. An argument in our system is a tuple of either the form [position, link, position] or [position, link, alternative] In other words, it links together a position with an alternative or another position. Our notion of argument differs from the above in that it does not presume that the support ....

Benferhat, S., Dubois, D., Prade, H.: How to infer from inconsistent beliefs without revising? In Proceedings of the 14th IJCAI, Montreal, 1995, pp. 1449-1455.

....have been proposed in the literature [28] in order to address various application areas [2] For instance, in [9] an argument consists of a support base, that may contain formulas which speak for or against a certain position, and a conclusion. A similar notion of argument has been given in [3] 2 . An argument in our system is a tuple of either the form (position; link; position) or (position; link; alternative) In other words, it links together a position with an alternative or another position belonging to a dioeerent issue. Agents put forward arguments to convince their opponents ....

Benferhat, S., Dubois, D., Prade, H.: How to infer from inconsistent beliefs without revising? In Proceedings of the 14th IJCAI, Montreal, 1995, pp. 1449-1455.

....arguments can simultaneously be applied. Various notions of an argument have been suggested in the literature. In [9] an argument consists of a support base, that may contain formulas which speak for or against a certain position, and a conclusion. A similar notion of argument has been given in [3]. All the above notions are extensions of the one proposed in [23] Our definition di#ers from the above in that it does not presume that the support base of an argument is minimal. That is, a strict subset of the support base of an argument can be the support base of another argument with the ....

Benferhat, S., Dubois, D., Prade, H.: How to infer from inconsistent beliefs without revising? In Proceedings of the 14th IJCAI, Montreal, 1995, pp. 1449-1455.

.... average method to perform S3 in MSBR. S4 consists of two substeps. Selecting a good CG from G. Normally, CG is the good with the highest priority in G . In case of ties, CG might be either one of those with the same highest priority (randomly selected) or their intersection (see [15] and [21]) This latter case means rejecting all the conflicting but equally credible information items. The result is not a good (it is not maximally consistent) and thus implies rejecting more assumptions than necessary to restore consistency. We believe that this could be avoided by simply considering ....

Benferhat S., Dubois D. and Prade H., How to infer from inconsistent beliefs without revising?, in Proc. of the 14th Inter. Joint Conf. on Artificial Intelligence, pp. 1449-1455, 1995.

....on a priority ordering. Then the handling of priorities has been shown to be completely in agreement with possibilistic logic (Dubois and Prade, 1991; Benferhat et al. 1992) One way of tackling inconsistency is to revise the knowledge base and restore consistency. However, as pointed out in (Benferhat et al. 1995), in the case of multiple sources of information, it does not always make sense to revise an inconsistent knowledge base since it comes down to destroying part of the knowledge. In the context of merging several knowledge bases, the introduction of priorities between pieces of information can be ....

Benferhat S., Dubois D., Prade H. (1995) How to infer from inconsistent beliefs without revising?. Proc. of the 14th Inter. Joint Conf. on Artificial Intelligence (IJCAI'95), Montréal, Canada, Aug. 20-25, 1449-1455.

No context found.

S. Benferhat, D. Dubois, and H. Prade. How to infer from inconsistent beliefs without revising. In Proc. of the 14 International Joint Conference on Artificial Intelligence (IJCAI'95), pages 1449--1455, Montreal, 1995.

No context found.

S Benferhat, D Dubois, and H Prade. How to infer from inconsistent beliefs without revising. In Proc. of the 14th Int. Joint Conf. on AI, 1995.

No context found.

S. Benferhat, D. Dubois, and H. Prade. How to infer from inconsistent beliefs without revising? In Proceedings of the Fourteenth International Joint Conferences on Artificial Intelligence, pages 1449--1455, 1995.

No context found.

S. Benferhat, D. Dubois and H. Prade. How to infer from inconsistent beliefs without revising? In Proceedings of the 14th InternationalJoint Conference of Artificial Intelligence. MorganKaufmann, 1995.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC